AMBER Archive (2007)

Subject: Re: AMBER: Ligand-Protein Distance Restraints during MD Simulation

From: David Mobley (dmobley_at_gmail.com)
Date: Tue Jan 30 2007 - 11:16:19 CST

Hi,

> Now, the starting complex of my GPCR-Ligand model has several
> H-bonding interactions including the above mentioned saltbridge/h-bond
> (resulting from: I built this model in vaccum and docked ligands as
> reported in PROTEINS 2006, 65, 824-842). Now, I am trying to simulate
> this system in membrane environment.
>
> Following are my doubts:
>
> Question 1A: After putting this above starting complex (with all
> h-bonds/salt-bridge) in a lipid bilayer, when I perform Molecular
> mechanics on this system, SHOULD I RESTRAIN THE PROTONATED NITROGEN
> AND CARBOXYLATE GROUP (invloved in saltbridge/h-bond), and also
> restrain other putative H-bonding interactions. DOES IT MAKE SENSE TO
> GIVE such DISTANCE RESTRAINTS (I have seen some publications which
> have used them during MD)

I know people sometimes do this. In my opinion it is essentially a
fudge to make things come out how you expect even if they won't
otherwise do so. In other words, if you run a simulation and the
hydrogen bonding interaction you expect doesn't remain stable, instead
of either (a) assessing whether you were wrong and the bond isn't
stable, or (b) trying to figure out if there are parameter problems,
you (c) FORCE the hydrogen bond to stay stable by adding restraints.

Obviously, I'm saying I don't think it makes much sense. It is not
clear to me what you can learn from a simulation where you force the
outcome you expect; in my mind the addition of the restraints calls
into question anything else you try to extract from the simulation
after the restraints are added (unless you use some sort of WHAM-like
reweighting scheme to try and unbias your results for the effect of
the restraints).

You are, of course, free to use restraints anyway -- and maybe people
disagree with me on this. But if I reviewed your paper I'd call you on
it.

> Question 1B: IF I SHOULD USE SUCH DISTANCE RESTRAINTS BETWEEN LIGAND
> AND PROTEIN ATOMS, THEN SHOULD THESE RESTRAINTS BE INCLUDED IN all
> 1,2,3 steps below or just certain of these steps (might sound like a
> stupid question, however, wanna make sure):
>
> 1. Initial Minimization of System
> 2. Equilibration MD
> 3. Production MD
>
> Ofcourse after my production MD, I will minimize the system without
> any restraints to check if the h-bonds etc. are stable??

If your h-bonds are stable during *minimization* without restraints,
you learn basically nothing, since minimization just takes you
downhill in energy. It would be *really* bad news if they weren't.

David

> I will look forward to your suggestions.
> Thanks and Best Regards
>
> Akshay
>
> = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
> Akshay Patny
> PhD Candidate (5th Yr.), Computational Chemistry
> Department of Medicinal Chemistry, School of Pharmacy
> The University of Mississippi
> 805 College Hill Rd, # 9, Oxford, MS 38655
> E-mail: akshay17_at_olemiss.edu
> Phone (O): (662)-915-1286,(M): (662)-801-5496
> = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu